Yoshua Bengio

In order to reduce the probability of someone intentionally or unintentionally bringing about a rogue AI, we need to increase governance and we should consider limiting access to the large-scale generalist AI systems that could be weaponized, which would mean that the code and neural net parameters would not be shared in open-source and some of the important engineering tricks to make them work would not be shared either. Ideally this would stay in the hands of neutral international organizations (think of a combination of IAEA and CERN for AI) that develop safe and beneficial AI systems that could also help us fight rogue AIs. (source)

I believe third-party audits should be mandatory for major AI systems. As these systems increasingly impact society, it's crucial to ensure they are safe, ethical, and unbiased through transparent and independent evaluations.

I believe it is crucial for big AI companies to allocate significant resources towards AI safety. Given the potential risks associated with advanced AI systems, dedicating a third of compute resources to ensuring safety can help mitigate unforeseen consequences and align AI development with human values.

As a leading figure in AI research, I strongly believe that waiting for a common agreement with the US and China before regulating AI would be a mistake. The EU should take proactive steps to implement ethical and robust AI regulations independently. Delaying such actions could hinder the protection of societal values, privacy, and intellectual rights, which are essential for responsible AI development.

I strongly believe that global AI governance frameworks must prioritize equitable access to AI technologies for all countries. AI has the potential to address many global challenges, but only if access is fair and inclusive. Developing nations should have the means to utilize AI to improve their economies and societies, and we must avoid a future where these transformative tools are concentrated in the hands of a few powerful nations.

I believe that no single lab should control the development of artificial superintelligence (ASI). Competition encourages innovation and ensures that multiple perspectives and approaches are explored, which is crucial for developing safe and ethical AI. An open and collaborative environment will better address global concerns and distribute the benefits of ASI more equitably.

I believe it's essential for governments to establish international AI ethics boards. The rapid development of AI technologies could have significant societal impacts, and international cooperation is crucial to ensure responsible AI deployment. By working together, we can establish guidelines and standards that prioritize ethical considerations and prevent potential harms.

As a scientist deeply involved in AI development, I believe we must be mindful of the broader implications of our work. It is irresponsible to pursue advancements at the risk of destabilizing international politics. Our focus should be on ensuring that AI contributes positively to society and global stability.

As an AI researcher, I believe that liberalism can be effective in a world with AI superintelligence. Safeguards and ethical frameworks must be implemented to ensure AI benefits humanity and upholds democratic values.